1. Field of the Invention
The present invention relates to a method and an apparatus for determining a projection area of an image.
2. Description of the Related Art
As computer technology and network technology develop, remote applications such as remote conferencing, remote medicine, etc., are realized and develop rapidly. In this field, the prior art has made certain progress. In this kind application, an image captured by an image pickup apparatus user is transmitted to a remote pickup apparatus; however, the remote pickup apparatus is not always interested in the image captured by the image pickup apparatus, and in fact, the areas in which the remote pickup apparatus user is interested often appear deformed because of imaging angle. Therefore, in the remote applications, it is always considered how to capture the areas in which the remote pickup apparatus user is interested and correct the deformation of the areas of interest. Therefore, processing load is increased, especially when it is required to accommodate the need of real-time.
In the cited reference No. 1 (U.S. Pat. No. 7,171,056B2), a method for converting the content of a white board to an electronic document is disclosed. For this purpose, in the cited reference No. 1, a method for detecting a static white board area is provided. However, this method is used only to detect the static white board, not realize the detection for real-time continuous projection areas.
In the cited reference No. 2 (U.S. Pat. No. 7,872,637B2), a system for following the location of a laser point in a projection image of a computer screen, comprising a step of detecting the projection area, is disclosed. However, in the cited reference No. 2, only space information from a single frame image is used to detect the projection area, and the focus is on path tracking for the laser point by performing a time analysis for locations of laser points in continuous images. In this technology, the process is performed independently for each frame without using the result of detecting the projection area from a previous frame; therefore, it makes the process of system processing burden heavy.
In the projection-pickup system, four corner points of the projection area are detected to correct the coordinates of the system. In the prior art, only the projection area of a single static image is detected, without performing a real-time detection by performing a space analysis for continuous images of a video stream. Furthermore, in the prior art, the projection area is determined by analyzing the whole image area, therefore computational complexity is increased.
Furthermore, in one implementation process of the remote application such as the remote conference, the projection area does not always remain constant, but the projection area may change for various reasons. In the prior art, the single frame image is processed independently, so the need of real-time processing of continuous video images is not considered or the need of real-time is met by the method of performing processing independently of each frame; therefore, it makes the system processing burden heavy. In the field of the remote application, it is seriously limited by the transmission condition of network, etc., therefore the problem of the heavy system processing burden is more serious than that of the local application.